Tears or Fears? Comparing Gender Stereotypes about Movie Preferences to Actual Preferences

This study investigated the accuracy of gender-specific stereotypes about movie-genre preferences for 17 genres. In Study 1, female and male participants rated the extent to which 17 movie genres are preferred by women or men. In Study 2, another sample of female and male participants rated their own preference for each genre. There were three notable results. First, Study 1 revealed the existence of gender stereotypes for the majority of genres (i.e., for 15 of 17 genres). Second, Study 2 revealed the existence of actual gender differences in preferences for the majority of genres (i.e., for 11 of 17 genres). Third, in order to assess the accuracy of gender stereotypes on movie preferences, we compared the results of both studies and found that the majority of gender stereotypes were accurate in direction, but inaccurate in size. In particular, the stereotypes overestimated actual gender differences for the majority of movie genres (i.e., 10 of 17). Practical and theoretical implications of these findings are discussed.

Introduction

Many people have gender stereotypes. Gender stereotypes are beliefs concerning the abilities, attitudes, preferences and behaviors of a ‘typical’ man or a ‘typical’ woman (Mealey, 2000). For example, many people believe that women have better verbal abilities than men, whereas men are assumed to have better mathematical skills than women. Whereas research has partly confirmed the former assumption (e.g., for verbal fluency, see Hyde and Linn, 1988), it has disconfirmed the latter (cf. Hyde, 2016, for a review). Because stereotypes may, at least partly, reflect real differences between groups, stereotypes are not necessarily inaccurate (see Hilton and von Hippel, 1996; Brown, 2010, for reviews on stereotypes in general). Hence, “the accuracy of stereotypes is an empirical question, not an ideological one” (Jussim et al., 2009, p. 199). Previous research has already addressed the accuracy of gender stereotypes concerning attitudes (e.g., Diekman et al., 2002) or cognitive abilities (e.g., Halpern et al., 2011; Hyde, 2016), but there is almost no research about the accuracy of gender stereotypes concerning media preferences. The present study thus compared presumed and actual differences in film preferences of men and women.

Gender and Movie Preferences

Everyday observations suggest the existence of strong and well-known stereotypes about gender differences in movie preferences, which can be summarized as follows. Women are supposed to like romantic and melodramatic movies, which are snidely called “chick flicks” or “tearjerkers”, as well as comedy, but to dislike action and horror movies. In contrast, men are supposed to like action and horror movies, but to dislike romantic and melodramatic movies. Empirical research on gender differences in actual movie preferences has confirmed, at least to some degree, these popular stereotypes (see Oliver, 2000; Greenwood and Lippmann, 2010, for reviews). The assumption that women prefer romantic and melodramatic movies as well as comedy, for instance, has received some empirical support (e.g., Oliver et al., 1998; Harris et al., 2000). The same is evident for the larger male than female preference for action and horror movies (e.g., Sparks, 1991; Krcmar and Kean, 2005).

Greenwood (2010), for example, investigated the impact of gender and mood state (happy vs. sad) on choosing between films from different genres. First, participants had to write about a happy or sad life event in order to induce the corresponding mood in these participants. Then, participants were asked to indicate which type of film they wished to see right now by using closed and open questions. The questionnaire required a choice between some happy film genres (e.g., romantic comedy, dark comedy), some sad film genres (e.g., social drama, romantic drama), action-adventure movies, and suspenseful movies (i.e., thriller). Results showed that, independent from induced mood, female participants preferred romantic genres (i.e., romantic comedy, romantic drama), whereas male participants preferred action-adventure movies and suspenseful films.

Further results showed that men, in stark contrast to women, have a preference for programs with violent or sexual content. For example, Blanchard et al. (1986) presented their participants violent material from different sources or genres (i.e., from a spy movie, a war movie, a western movie, and a cartoon). When interviewed afterwards, male participants reported more pleasure from viewing the violent films than the female participants (see also Koukounas and McCabe, 2001). A stronger male than female preference was also found for horror movies that typically contain frightening and violent scenes (e.g., Cantor and Reilly, 1982; Zillmann et al., 1986; Oliver, 1993).

Finally, strong gender differences have also been observed in the use and reception of pornographic material. A couple of studies revealed that men consume pornographic films and materials more frequently than women (e.g., Bryant and Brown, 1989; Hald, 2006; see Peter and Valkenburg, 2016, for reviews). Moreover, men repeatedly reported stronger affective and physiological responses to pornographic films than did women (e.g., Janssen et al., 2003). However, the size of the difference between male and female responses to pornographic material has been found to vary with some characteristics of the films (e.g., Murnen and Stockton, 1997; Janssen et al., 2003; Hald and Štulhofer, 2015).

To summarize, previous studies on film and movie genre preferences have already discovered some reliable gender differences, as described above. However, most of these studies involved relatively small numbers of movie genres (i.e., 2–4; e.g., Sparks, 1991; Oliver et al., 1998), and have focused on a relatively small subset of genres, including action, drama, horror, sports, and pornography. Moreover, in most studies, participants judged parts of individual films or verbal descriptions of fictive films, and the representativeness of these materials constrains any attempts to generalize the results to the genre in question. For these reasons, we do not yet know much about some movie genres (e.g., fantasy films, history films, mystery films, science-fiction films, crime movies, western films and, as a typical German genre, ‘heimat’ films).1 And we do not yet know much about the relative popularity of movie genres within and across both sexes, either. Finally, studies typically addressed the actual preferences of men and women for movie genres, whereas gender stereotypes in movie preferences have neither been studied nor have stereotypes been compared to the actual movie preferences of men and women.

Possible Sources of Gender Stereotypes

Although the data presented here are not suitable for answering the question of where gender stereotypes and gender differences with respect to media preferences come from, it might be worthwhile to summarize the most prominent accounts of gender differences in movie preferences. These accounts can be classified in four groups: theories referring to media content, biological accounts, evolutionary accounts, and accounts in terms of gender socialization (e.g., Oliver, 2000; Hust and Brown, 2008).

A first group of accounts refers to differences in the content of films from different genres, such as the main topic of the film or the protagonists’ gender. Regarding film topics, it was suggested that women prefer movies or genres centering on social relationships, whereas men prefer movies or genres centering on aggressive conflicts (e.g., Oliver et al., 2000). Regarding the gender of film protagonists, it was suggested that observers especially enjoy films if they can identify with a major protagonist, and observers prefer to identify with same-sex protagonists (e.g., Hoffner and Cantor, 1991; Hoffner, 1996). In fact, content analyses showed that male protagonists dominate in film genres that are preferred by men (e.g., action films, ‘male’ sports), whereas female protagonists are more likely to be found in film genres that are preferred by women (e.g., daily soaps, melodramatic movies).

Biological accounts refer to biological differences between men and women (e.g., Malamuth, 1996; Goldstein, 1998). For example, the male hormone testosterone is related to sex drive and dominant behavior in men (e.g., Mazur and Booth, 1998). This relationship could provide an account for the stronger interest of men, compared to women, in movies featuring competition and sex. The female hormones oxytocin and vasopressin, however, are supposed to be related to pair bonding and feelings of love in humans (e.g., Fisher, 1998; Carter, 2007). From a neurobiological perspective this difference could be responsible for women’s greater interest in romantic movies. It should be noted though that biological accounts are controversially discussed. For example, researchers from the field of “gendered neurocultures” argue that biological accounts should critically be reflected in terms of gendered ascriptions that are linked to biological processes (e.g., Fillod, 2014; Matusall, 2014).

Evolutionary accounts describe how human evolution may have shaped gender differences in media preferences. For example, from parental investment theory (Trivers, 1972), it can be predicted that “women prefer media contents dealing with issues of mate choice, partner loyalty, and the loss of a partner … [and] men are more likely to select media contents dealing with kin protection, rivalry, status, power, and the acquisition and maintenance of resources” (Hennighausen and Schwab, 2015, p. 144). In a sample of movie visitors, Schwab (2010) replicated the finding that women prefer melodramatic and romantic movies, whereas men prefer action movies. Interestingly, and in some contrast to previous findings, Schwab observed no significant gender differences for other genres (i.e., comedy, horror, and thriller).

A fourth group of accounts refers to differences in the socialization of men and women and, as a result, to differences in learned gender roles. Two important theories of gender socialization are social-learning theory and gender-schema theory. According to social learning theory, children learn the gender role matching their biological sex by operant conditioning and/or by latent learning (e.g., Mischel, 1966). Hence, the agents of socialization (e.g., parents, peers, and teachers) are assumed to reward children for consuming gender-role congruent media content, and to punish (or, at least, refrain from rewarding) children for consuming gender-role incongruent media content. In addition, children may also learn gendered media preferences by observing and imitating same-sex role models. Whereas social-learning theory focuses on the acquisition of gendered behavior, gender-schema theory focuses on the cognitive representation of the gender role. A gender schema is a cognitive structure representing knowledge about the (typical) features of a gender (e.g., Bem, 1981). During socialization, children acquire gender schemas and form associations between the contents of their self-concept and the contents of the gender schema matching their biological sex. Importantly, the gender schema not only represents knowledge about gender, but it also biases the processing of information towards the own gender. Hence, on this account, boys and girls develop different media preferences because (different) gender schemas direct their attention towards gender-congruent media contents.

Presumably, not a single account is able to explain gender differences in media preferences in their entirety. More likely, a combination of different accounts provides the most valid explanation. For example, it might be that men and women are born with some innate biological cognitive mechanisms that are the result of human evolution. Those mechanisms might then be further shaped by environmental factors and thus, finally, cause the observed preferences, for instance for a specific movie content.

Accuracy of Gender Stereotypes

Only recently, researchers have begun to empirically assess the accuracy of social stereotypes, and to compare stereotypes about groups and group differences to measurements of group attributes and group differences (see Ryan, 2002; Jussim et al., 2009, for reviews). The majority of these studies revealed that people are able to provide accurate judgments for many attributes of men and women (Hall and Carter, 1999; Jussim et al., 2009). For example, Swim (1994) compared stereotypes about gender differences for 17 characteristics (e.g., aggression, cognitive abilities, non-verbal behaviors) to measures of actual gender differences as derived from meta-analytic studies. She observed accurate judgments of gender differences for the majority of characteristics under investigation. The only overestimations of gender differences occurred for aggression and verbal abilities. Similarly, Diekman et al. (2002) reported that, on average, male and female participants quite accurately predicted the political and social attitudes of men and women.

The evidence on the accuracy of people’s beliefs about gender differences in cognitive abilities appears somewhat mixed, however. Beyer (1999), for example, found that university students misperceived the direction of the gender difference in performance (i.e., the grade point average, GPA) for a set of ‘masculine’ majors2: Participants believed that male students perform better in these majors, although female students actually achieved higher grades. As already stated, Swim (1994) reported that participants quite accurately judged (existing) gender differences in math performance, but overestimated gender differences in tests on verbal abilities. Finally, Halpern et al. (2011) found that participants were generally accurate about the direction of gender differences in the cognitive domain, but underestimated the actual size of the differences. In their recent review on the accuracy of gender stereotypes, Jussim et al. (2009, p. 211) conclude that, “at least a plurality of judgments was accurate” and that “there was no support for the hypothesis that stereotypes generally lead people to exaggerate real differences.”

Determining the accuracy of stereotypes in general and of gender stereotypes in particular is important for at least three reasons. First, assessing and analyzing the accuracy of stereotypes may provide important clues on the origins and development of stereotypes. For example, Hall and Carter (1999) identified several abilities and traits (e.g., interpersonal sensitivity) that were correlated with the accuracy of gender stereotypes. Second, the evidence demonstrating that a stereotype is wrong could (and should) be used to correct this stereotype and to improve the relationship between social groups (e.g., Beyer, 1999). Third, comparing presumed preferences (i.e., stereotypes) to the actual preferences of social groups may provide helpful information for some industries, including the movie industry in case movie preferences are concerned. In particular, knowledge about discrepancies between presumed and actual gender differences in movie preferences may be helpful for addressing a broader audience. If, for example, the actual gender gap in preference for a particular movie genre is larger than presumed, one might reduce this difference by producing more gender-neutral movies of this genre. Also, movie producers might benefit from knowing the actual size of the gender gap by focusing more on producing gender-typical movies.

The Present Study

The present investigation had four aims. The first aim was to investigate—for the first time (at least to our knowledge)—the stereotypes of men and women about the movie preferences of men and women (Study 1). The second aim was to investigate gender differences in stereotypes about the movie preferences of men and women (Study 1). In other words, we wanted to investigate whether men and women have similar or different stereotypes about gender-specific movie preferences. The third aim of our study was to investigate the actual preferences of young adults for our set of movie genres (Study 2). Hence, we addressed more genres than most of the previous studies, and included genres that have not been investigated (often) before (e.g., ‘heimat’, history, mystery, crime, thriller, and western). Finally, the fourth aim of our study was assessing the accuracy of gender stereotypes concerning movie preferences by comparing the results of Study 1 to those of Study 2.

We assessed and compared presumed gender differences in movie preferences to actual gender differences in movie preferences for 17 movie genres. The term “genre” refers to a category of motion pictures that are similar with regard to particular features, including topic (e.g., adventure, war), geographical or historical background (e.g., Western, science fiction), stylistic issues (e.g., animation) and the targeted audience (e.g., children vs. adults; cf. Grant, 2003). For our investigation, we chose 17 well-known movie genres3: (1) action movie, (2) adventure movie, (3) animation movie, (4) comedy movie, (5) crime movie, (6) drama, (7) erotic movie, (8) fantasy movie, (9) ‘heimat’ film, (10) historic movie, (11) horror movie, (12) mystery movie, (13) romance, (14) science fiction movie, (15) thriller movie, (16) war movie, (17) Western movie. Only the genre names were used in both studies. We did not present particular examples for each genre because we wanted to investigate how participants think about movie genres, and not how they think about particular movies, and we had several additional reasons for doing that. First, we were sure that our participants would recognize each of the 17 movie genres used in our study without providing examples. Second, the representativeness of a particular movie for a genre is always debatable. Third, example movies might be familiar to some participants, but not to all. Hence, providing no examples created equal conditions for all participants. Finally, we deliberately chose genres of movies that are typically presented in the cinema (and on TV). Hence, we excluded common TV formats like daily soaps, documentaries, sports, talk shows, sitcoms and so on.

Study 1: Stereotypes About Movie Preferences of Men and Women

In Study 1, we investigated the stereotypes of young adults about movie genre preferences of men and women in order to address two research questions. First, we asked which genres are generally associated with men and which genres are generally associated with women. For some genres, we expected to observe the popular gender stereotype. In particular, we expected that a majority of participants would ascribe romantic movies and drama to women. Similarly, we expected that a majority of participants would ascribe more action-packed and violent genres (i.e., action, adventure, horror, thriller, and war movies) to men. We had no predictions regarding the remaining genres. Our second research question asked whether male and female participants would express similar or different views (i.e., stereotypes) about the movie preferences of men and women.

Materials and Methods

Ethics, Consent, and Permissions

The research reported in our manuscript meets the ethical guidelines of the German Society of Psychology (Deutsche Gesellschaft für Psychologie, DGPs) and is consistent with the principles of research ethics as published by the American Psychological Association (APA). Data collection was completely anonymous. That is, except for age and gender, we did not record personal information from the participants. All participants in the two studies reported here gave informed consent before filling out the questionnaires on movie preferences. The research reported here had not been approved by a local ethics committee because the ethical guidelines of the German Society of Psychology do not require ethical approval of basic psychological studies involving simple behavioral data.

Participants and Interviewer

One hundred and fifty volunteers participated in Study 1. The sample consisted of 75 women (mean age = 23.1 years; SD = 4.0) and 75 men (mean age = 24.1 years, SD = 3.8). The large majority of participants were students (undergraduates) at a public university in Germany. The interviewers were 15 students (12 female, 3 male), and each interviewed 5 men and 5 women.

Materials

We developed a questionnaire with 24 items4. Items 1–4 asked for some demographic characteristics of our participants (gender, age, main profession or major study course, minor study course). Then, items 5–21 asked the participant to indicate for each of 17 movie genres the degree to which the respective genre is, in her or his opinion, preferred by men or women. In particular, each item asked: “What do you think: Are (adventure) films preferred by men or by women?” Participants had to indicate their answer on an 11-point rating scale. The ticks were numbered from 5 to 0 and to 5 again. The central tick (0) was always marked as “equally preferred by men and women”. For half of the questionnaires (i.e., participants) the leftmost tick was marked as “exclusively preferred by women”, whereas the rightmost tick was marked as “exclusively preferred by men”. For the other half of questionnaires (i.e., participants) the leftmost tick was marked as “exclusively preferred by men”, whereas the rightmost tick was marked as “exclusively preferred by women”. The items 5–21 probed for the following movie genres (the German terms are given in brackets): adventure [Abenteuerfilm], action [Action-Film], animation [Animationsfilm], comedy [Filmkomödie], crime [Kriminalfilm], drama [Filmdrama], erotic [Erotikfilm], fantasy [Fantasy-Film], ‘heimat’ [Heimatfilm], history [Historienfilm], horror [Horrorfilm], romance [Liebesfilm], mystery [Mystery-Film], science fiction [Science-Fiction-Film], thriller [Thriller], war [Kriegsfilm], and Western [Westernfilm]. The items 22–24 asked our participants for some quantitative aspects of TV and film consumption. Item 22 asked how many hours per day, on average, the participant watched TV. Item 23 asked how many films, on average, the participant consumed per week. Finally, item 24 asked how often, on average, the participant went to the cinema per month.

We constructed four versions of our questionnaire in order to minimize (a) the effects of marker positions and (b) the effects of item order. As already said above, in half of the questionnaires the left pole of the rating scales referred to women, whereas the right pole referred to men, and the reverse was true for the other half of questionnaires. This manipulation was crossed with two different orders of movie genres. For one half of the questionnaires, items 5–21 were jumbled to produce a random order of genres. This (random) order was reversed for the other half of the questionnaires. We made sure that similar numbers of men and women filled out each version of the questionnaire.

Procedure

People were asked to participate in a study on the film preferences of men and women. Participants filled out the questionnaire on themselves, in order to minimize interviewer effects. If necessary, the interviewer provided a pen and a writing pad. Data were always collected in dyads of one interviewer and one participant. Participants took either part for free or received course credits.

Read all articles from this link https://www.frontiersin.org/articles/10.3389/fpsyg.2017.00428/full